arX

iv:1

602.

0179

7v1

[as

tro-

ph.S

R]

4 F

eb 2

016

Resolved Companions of Cepheids: Testing the Candidates with

X-Ray Observations1

Nancy Remage Evans,1 Ignazio Pillitteri,1,2 Scott Wolk,1 Margarita Karovska,1 Evan

Tingle,1 Edward Guinan,3 Scott Engle,3 Howard E. Bond,4,5 Gail H. Schaefer,6 and Brian

D. Mason7

Received ; accepted

1Smithsonian Astrophysical Observatory, MS 4, 60 Garden St., Cambridge, MA 02138;

nevans@cfa.harvard.edu

2INAF-Osservatorio di Palermo, Piazza del Parlamento 1, I-90134 Palermo, Italy

3Department of Astronomy and Astrophysics, Villanova University, 800 Lancaster Ave.,

Villanova, PA, 19085, USA

4Dept. of Astronomy & Astrophysics, Pennsylvania State University, University Park,

PA 16802; heb11@psu.edu

5Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218

6The CHARA Array of Georgia State University, Mount Wilson, California 91023, USA;

schaefer@chara-array.org

7US Naval Observatory, 3450 Massachusetts Ave., NW, Washington, DC 20392-5420

– 2 –

ABSTRACT

We have made XMM-Newton observations of 14 Galactic Cepheids that have

candidate resolved (≥5′′) companion stars based on our earlier HST WFC3 imag-

ing survey. Main-sequence stars that are young enough to be physical companions

of Cepheids are expected to be strong X-ray producers in contrast to field stars.

XMM-Newton exposures were set to detect essentially all companions hotter than

spectral type M0 (corresponding to 0.5 M⊙.) The large majority of our candi-

date companions were not detected in X-rays, and hence are not confirmed as

young companions. One resolved candidate (S Nor #4) was unambiguously de-

tected, but the Cepheid is a member of a populous cluster. For this reason, it

is likely that S Nor #4 is a cluster member rather than a gravitationally bound

companion. Two further Cepheids (S Mus and R Cru) have X-ray emission that

might be produced by either the Cepheid or the candidate resolved companion.

A subsequent Chandra observation of S Mus shows that the X-rays are at the

location of the Cepheid/spectroscopic binary. R Cru and also V659 Cen (also

X-ray bright) have possible companions closer than 5′′ (the limit for this study)

which are the likely source of X-rays. One final X-ray detection (V473 Lyr) has

no known optical companion, so the prime suspect is the Cepheid itself. It is a

unique Cepheid with a variable amplitude.

The 14 stars that we observed with XMM constitute 36% of the 39 Cepheids

found to have candidate companions in our HST/WFC3 optical survey. No young

probable binary companions were found with separations of ≥5′′ or 4000 AU.

Subject headings: stars: binaries — stars: massive — stars: formation — stars:

variable: Cepheids

– 3 –

1. Introduction

Binary/multiple configurations influence every phase of a star’s life: formation, the

main sequence, and the post-main sequence, and often drastically affect the outcomes. For

instance, Kraus et al. (2012) find that binaries with intermediate separations (40 AU) are

less likely to have protoplanetary disks than both closer and wider binaries, and hence are

less likely to form planets. For main-sequence stars, Sana et al. (2012) recently showed

that more than 70% of O stars (≥20M⊙) are destined to undergo binary interactions with

mass exchange, and a third of those undergo mergers. At later stages, many classes of

X-ray sources have scenarios involving binaries containing a compact object (supernovae

progenitors, X-ray binaries, novae, cataclysmic variables, and symbiotic stars). However,

there are many unanswered questions about how these configurations are achieved.

Technical developments in recent years in both radial velocities and imaging have

contributed greatly to the studies of binary/multiple stars. For massive stars, the velocity

study of O stars in the 30 Dor (Tarantula Nebula) in the LMC (Sana et al. 2013) is of

particular importance. It confirmed the large fraction of binaries, and demonstrates the

importance of interaction inat the their evolution (above; Sana et al. 2012). A comparable

velocity survey was done on the B stars in the Tarantula Nebula (Dunstall et al. 2015),

who concluded that the multiplicity properties of the B stars are essentially the same as for

the O stars. An extensive high-angular-resolution study of O stars was made at the Very

Large Telescope (VLT) and the Very Large Telescope Interferometer (VLTI) which covers

separations from 0.5 to 104 mas (1 to 20,000 AU at a typical distance of 2 kpc; Sana et

al. 2014). Of particular importance was filling the gap between spectroscopic binaries and

visual binaries which had been found by Mason et al. (1998). A further survey was made

1Based on observations obtained with XMM-Newton, an ESA science mission with in-

struments and contributions directly funded by ESA Member States and the USA (NASA).

– 4 –

by Aldoretto et al. (2015), of O and B stars using the HST Fine Guidance Sensor (FGS). A

summary of previous studies of binarity/multiplicity in massive stars can be found in Evans

et al. (2013) and Evans et al. (2016a).

Binary/multiple properties reflect the processes of star formation in their extent

and mass ratios. The formation of wide binaries (1000 AU or greater) is a problem for

star formation, since this is larger than the typical size of a star-forming core. Several

mechanisms have been suggested to get around this problem. One is the “unfolding” of

a triple system. In a triple which is dynamically unstable, one component (typically the

smallest) may be sent to a wider orbit (Reipurth & Mikkola 2012). Another mechanism is

the acquisition of a third star during cluster dissolution (Kouwenhoven et al. 2010). Parker

& Meyer (2014) concluded that over time the reduction of triple systems through disruption

is balanced by acquisition of another star, keeping the binary fraction relatively constant.

The stars targeted in the present study are wide resolved companions. Within

such systems there is ample room for a close binary, and hence wide companions are

candidates for triple systems. In the same way that early exoplanet discoveries taught us

the importance of migration, analysis of multiple systems points to dynamical evolution in

which a third star plays a special role. One example of the consequences of a triple system

is that spectroscopic binaries with P < 3 d are much more likely to have a third component

than those with P > 12 d (Tokovinin et al. 2006). This implies that dynamical evolution has

shortened the periods of the first group. Conservation of total system angular momentum

requires that the wider tertiary components move to wider separations. Scenarios involving

a triple system have also been proposed to explain close eccentric binaries (Perets & Kratter

2012) and close binaries with small mass ratios (Moe & Di Stefano 2015), both required for

exotic end-stage objects.

The present study is Paper IV in a series aimed at determining the binary/multiple

– 5 –

properties of Cepheids. They are reasonably massive stars (typically 4 to 8M⊙ stars,

abbreviated to 6M⊙ below) in a well-known evolutionary state with well-studied distances

and reddenings. This study builds on a Hubble Space Telescope (HST) Wide Field Camera

3 (WFC3) imaging survey of 70 Cepheids. Evans et al. (2013; Paper I) demonstrated the

technique using Cepheids with reasonably massive companions (q = M2/M1 ≥ 0.4). In

Papers II and III (Evans et al. 2016a and 2016b), we used the WFC3 images in the F621M

and F845M filters to identify candidate resolved companions as close as approximately

0.′′5. With this we can identify main-sequence companions typically as close as 300 AU.

Paper II discusses possible companions ≥ 5” from the Cepheid; Paper III discusses closer

companions identified after point spread function (psf) correction.

Possible resolved companions were identified in Paper II as follows. A color-magnitude

(CMD) was formed for stars from the F621M and (F621M–F845M) data, which transform

well to V and (V-I). This was compared with an isochrone at the distance and with the

reddening of the Cepheid (see Paper II). Stars in the field within 2 σ of the isochrone with

colors hotter than M stars were considered candidate resolved companions, and are listed

in Paper II.

Identification of resolved companions, particularly low-mass stars, is plagued by

contamination from galactic field stars. However, resolved companions which are genuine

physical companions of Cepheids (typically 50 Myr old) will be young. Stars this young

are active X-ray producers as compared with old field stars. In the present study, we

have observed a number of possible resolved companions from the HST survey with the

XMM-Newton X-ray satellite (XMM ) to identify X-ray active young stars, likely to be

genuine bound companions.

Thus, the aim of this study is to identify the widest companions of Cepheids down to

very small masses. One goal is to determine the maximum separation which set by star

– 6 –

formation plus subsequent dynamical evolution. A second aim is related to the fact that

a wide system may easily contain a closer binary, at least in the main-sequence phase. A

wide companion certainly does not guarantee a three-star system, but it is a flag that this

may be the case. As discussed above this opens a number of possibilities for its evolution.

The subsequent sections in this paper discuss the observations and data analysis (§2),

the resolved companions, upper limits and detections (§3), results (§4), discussion (§5), and

a summary.

2. Observations and Data Analysis

Our XMM target list was selected from a compilation of possible resolved companions

found in our HST optical imaging survey. Discussion of the results of the survey is

done in two parts, determined by data-reduction challenges (Papers II and III). Possible

companions at ≥5′′ are good targets for XMM imaging, since this is approximately the size

of the XMM psf. Specifically, our target list was drawn from Table A1 in Paper II of such

candidates, which would be resolvable by XMM. Table 1 lists the observed Cepheids, the

date of observation and UT at the beginning of the integration, the exposure duration for

the MOS1 CCD camera, the reddening and distance (from Paper II), and the XMM filter

used. In addition to XMM observations made for this study, two observations were obtained

from the XMM archive: U Sgr (PI: Motch) and ℓ Car (PI: Guinan). Of the 39 Cepheids

with candidate resolved companions ≥5′′ (from the 70 WFC3 observations), altogether 14

of them (36%) were observed with XMM.

Data analysis was carried out as in Pillitteri et al. (2013), using standard XMM data

reduction (SAS) tasks to filter events so that events in the band 0.3 to 8.0 keV were

used. Only good time intervals were included and high-background intervals were removed.

– 7 –

Source detection and upper-limit calculation were done using a wavelet deconvolution

algorithm, as implemented in the code originally written for ROSAT images (Damiani et

al. 1997a, 1997b) and adapted for XMM images.

3. Candidate Resolved Companions

The results of our XMM observations of possible resolved companions are summarized

in Table 2. The successive columns are: the companion identification number for each

Cepheid, V and V − I from Paper II, the separation in arcsec, the position angle, the

separation converted to AU using the distance from Table 1, the J2000 RA and Dec of

the companion, and whether an X-ray source was detected at that location (Y/N/?). The

coordinates were measured from the WFC3 images.

In only three cases (S Mus, R Cru, and S Nor #4) is there possible X-ray flux at the

location of the resolved companion, as shown in Fig 1, Fig 2, and Fig 3. In two further

cases (V659 Cen and V473 Lyr) a source was detected at a position indistinguishable from

the Cepheid itself. These will be discussed in §3.2.

3.1. Detection Upper Limits

The XMM exposure times were set with the aim of detecting main-sequence companions

as cool as spectral type K, which would have logLX of 29.2 erg sec−1 or greater (at the

distance and with the reddening of each Cepheid). X-ray fluxes for M dwarfs become much

fainter as temperature decreases, and exposure times become much longer. X-ray flux, of

course, depends on the age of the stars. The α Per cluster is approximately 50 Myr old,

making it an appropriate comparison for Cepheids which have about the same age. ROSAT

observations of the cluster were discussed by Randich et al. (1996). At approximately this

– 8 –

limit, they detected 88% of K stars in their list. Pillitteri et al. (2003) use ROSAT data to

obtain the X-ray luminosity distributions of G, K, and M stars for the α Per cluster. The

X-ray detection rate at logLX of 29.2 erg sec−1 is 80-90% for G and K stars, but falls off

for M dwarfs. A deeper XMM observation of the α Per cluster was discussed by Pillitteri

et al. (2013). Their luminosity function found a similar detection rate. (Note that the

relevant bin in there discussion includes early M stars as well as G and K stars). Because

of the dependence of X-ray flux on both age and spectral type, and, of course, the X-ray

variability cycles of late-type stars, there is some imprecision in the estimate. However, this

luminosity limit should ensure that most late-type stars through the K range (to 0.5 M⊙)

will be detected.

To quantify the non-detections in Table 2, we have computed upper limits at the

positions of each of the candidate companions. Since we know the location of the possible

companion stars, we used a 3σ detection limit for the upper limit. We ran the wavelet

algorithm with this threshold to determine the level of count rates for a point-like source

at the positions of the Cepheid companions. Table 3 (Col. 3) lists the upper limit. The

detections and upper limits were obtained using both MOS and pn images, using the

standard relative efficiency MOS/pn cameras of 800/260 from the effective areas in the

appropriate energy range. For V473 Lyr the count-rate limit is measured slightly further

from the Cepheid than the position shown in Fig. 5, to decrease flux from the position of

the Cepheid (although there is still likely to be some contribution from the source at the

Cepheid position). The ratio of NH to E(B − V ) used to derive the column densities in

Col. 4 was 5.9 × 1021cm2mag−1 . The corresponding flux (Col. 5) was generated for this

NH using an APEC plasma emission code in the PIMMS flux calculation software 2 with

log T = 7.0 and solar abundance for XMM using the filter in Table 1. The upper-limit

2http://cxc.harvard.edu/toolkit/pimms.jsp

– 9 –

X-ray luminosity (Col. 6) was then derived using the distance.

In summary, most of the upper limits (Table 3) are close to the goal of logLX of 29.2

erg s−1, showing that we should have detected most of the late F, G, and K stars which are

young and at the distance of the Cepheids.

3.2. X-ray Detections

The candidate optical companions targeted by our XMM observations are ≥5′′ from

the Cepheid, allowing us to avoid confusion with X-ray emission from the Cepheid itself.

In five cases, we detected an X-ray source of which four are at or near the Cepheid itself.

These detections are listed in Table 4, which gives fluxes, luminosities, and comments about

the sources. The detections of each source are discussed below. For comparison, several

Cepheids (Polaris, δ Cep, and β Dor) have been detected in X rays with typical luminosities

of logLX = 28.6–29.0 erg s−1. The XMM observation of β Dor has a logLX of 29.0 ergs

s−1, and a soft spectrum (Engle et al. 2009; Engle et al. 2014). Polaris was observed with

Chandra (Evans et al. 2010) and has logLX = 28.9 erg s−1, and also a soft spectrum. As

they discussed, the interpretation is complicated by the fact that Polaris has a low-mass

companion which could also produce or contribute to the X-ray flux. δ Cep itself has

XMM observations at several phases. The flux is reasonably constant at a luminosity of

about logLX = 28.6 erg s−1. One point is significantly brighter, suggesting variability with

phase, which would indicate that the flux is produced by the Cepheid. However, recently

high-accuracy radial velocities show low-level orbital motion, consistent with a low-mass

companion (Anderson et al. 2015). As with Polaris, the companion could contribute some

or all the X-ray flux. The luminosities of these three systems provide either detections or

upper limits to the X-ray luminosity of the Cepheid itself, which is below the luminosities

in Table 4

– 10 –

In this section we discuss each of the XMM detections in Table 4.

S Mus: A possible companion of S Mus lies within the psf of the XMM image

(Fig 1). From the image we cannot tell whether the X-rays come from the companion

or from the Cepheid itself. We have recently obtained an observation of S Mus with the

Chandra X-ray satellite. From the image showing that the X-rays are not produced by the

resolved companion, but from the location of the Cepheid. Full discussion of the results is

in preparation. S Mus is a well-known binary system with a period of 1.38 years (Evans et

al. 2006, and references therein). Velocities of the hot companion have been measured in

the ultraviolet (Bohm-Vitense et al. 1997) and show no sign that the companion is itself a

binary. The companion has a spectral type of B3 V, and is the one companion among the

XMM observations hot enough that it could produce X-rays itself, which seems the most

likely source of the X-rays.

R Cru: As with S Mus, the XMM image (Fig 2) cannot determine whether the X-rays

come from the close (8′′) companion or the Cepheid. In addition, in Paper III, we find a

possible companion 1.′′9 from the Cepheid. Either of the companions or the Cepheid could

produce the X-rays. Fig. 2 also shows that there is a small brightening in X-rays at the

position of the resolved companion to the right of the Cepheid. We have examined that

area carefully, and there is not a significant source at that location, but rather a fluctuation

similar to other background fluctuations apparent in the figure.

V659 Cen: None of the five resolved possible companions in Fig 4 are coincident with

an X-ray source. However, in Paper III, there is a possible companion at about 0.′′7 from

the Cepheid. Either it or the Cepheid could produce the detection.

V473 Lyr: The X-rays in Fig 5 come from the vicinity of the Cepheid rather than

the possible companion (separated by 15′′). This raises the question whether there is a

closer low-mass companion which could produce the X-rays. There are considerable data

– 11 –

which address this topic. In Paper III there is no sign of a close companion. Conclusions

from velocities are tricky because of the variable pulsation amplitude. The star has been

discussed recently by Molnar & Szabados (2014), who conclude that it does not show

binary motion, based largely on the 5 years of velocity data from Burki (1984). A binary

companion can never be completely ruled out, but there is no evidence for one from either

velocities or imaging. Thus, the Cepheid itself is a possible source. This star is one of the

most unusual Cepheids known, unique among classical Cepheids in having a large-scale

amplitude variation, resembling the Blazhko variations in RR Lyrae stars.

S Nor #4: This is the only X-ray source that is unequivocally from the resolved

companion (Fig 3). However, since the Cepheid is a member of a populous open cluster,

there is a significant chance of an alignment with a cluster star.

In Table 4 we include the median energy for each of the detections. The sources on the

images are relatively weak, so full spectral analysis is not feasible. However, the median

energy provides information which can help distinguish between a low-mass star and a

cool supergiant, which has a softer spectrum. As an example, see Fig. 3 in Evans et al.

(2010), which compares the Polaris spectrum with the stacked spectrum from Orion Nebula

low-mass stars (Feigelson et al. 2005). The low-mass stars have a sharp energy peak at 1

keV, with a tail extending out to 2 keV. Stars with earlier spectral type (as well as older

stars), including the Polaris system, have energy peaks at about 0.9 keV.

The combination of the median energy values and LX in Table 4 provides the following

information about the X-ray sources. S Nor #4 is the only undisputed X-ray source resolved

from the Cepheid. Its median energy (0.95 keV) and LX are consistent with a low-mass

star. In fact, the LX for all the detected sources is consistent with low-mass stars, but S

Mus is the outlyer in the high end. For the four systems where the position of the X-ray

source is indistinguishable from the Cepheid, R Cru is the star with the highest median

– 12 –

energy (1.24 keV). This is a strong indication that the X-rays are produced by the close

(1.′′9) late-type companion with a harder spectrum than a supergiant. On the other hand,

the much softer median energy for V473 Lyr points to either the Cepheid itself or an as yet

unknown companion earlier than G spectral type. For V659 Cen a low-mass companion

is consistent with both the median energy and LX , and is there other evidence of such a

companion. For S Mus, the median energy and LX are consistent with X-rays produced by

shocks in an outflowing wind, as might be produced in the hot companion (see, for example,

Stelzer, et al. 2005)

Only the S Mus source has enough counts to make it reasonable to look at the spectrum

(Fig. 6). Although the counts warrant only crude energy bins, energy maximum is clear.

The resulting energy from the fit (kT = 0.64± 0.14 keV) is consistent with a low-mass star,

although the NH ((5± 2)× 1021 cm−2) and hence the unabsorbed flux (6.3× 10−14 erg cm−2

s−1 in the 0.3-8.0 keV band) is larger than that in Table 4.

In summary, of the five XMM detections, two stars (R Cru and V659 Cen) have a close

companion (<2′′ from the Cepheid) which are the most likely sources of the X-rays. For S

Mus), the working model is that the X-rays are probably produced by the B3 companion.

For S Nor, the late-type resolved companion is detected, though whether it is gravitationally

bound to the Cepheid or a cluster member is unknown. For the final detection, V473 Lyr,

the only candidate for the X-ray source appears to be the Cepheid.

We stress that the main goal of this study was to search for X-rays from young

low-mass stars to identify Cepheid companions. These are coronal stars, for which the

X-ray properties have been well studied. The mechanism for producing X-rays in Cepheids

themselves is not well understood. They are also coronal stars. One possible mechanism

is pulsation-driven shocks, as discussed by Engle, et al. (2014). Another possibility is

collisionless shocks (Ruby, et al. 2016).

– 13 –

4. Results

The distribution of separations for the possible companions investigated in Table 2

is shown in Fig 7. Appropriate to XMM’s spatial resolution, the Cepheids observed all

have possible companions at separations ≥5′′. In Fig 7, the optically resolved candidate

companions are overwhelmingly not found to have the X-ray strength expected for a

low-mass star about 50 Myr old. The star which is the exception to the finding that there

are no companions in the survey at greater distances than 8000 AU is the companion to

S Nor. This is one of two stars in the survey (S Nor and U Sgr) which is a member of a

populous open cluster. This increases the likelihood of a chance alignment with a cluster

member. We thus consider the S Nor #4 companion at a projected separation of 13,300

AU, to be an outlier, and most likely not to be gravitationally bound to the Cepheid.

The HST WFC3 survey covered 70 Cepheids, of which 39 were found to have

possible resolved companions with separations ≥5′′. We observed 14 (36%) of the possible

companions with XMM and found no young stars (with the exception of S Nor #4, which

is most probably a cluster member). Thus, we have found no probable resolved companions

with separations larger than 3950 AU (S Mus, Table 2).

5. Discussion

Systems containing Cepheids have undergone several periods of reorganization. As

massive stars they are highly likely to be found in binary or multiple systems. Much of the

dynamical adjustment from the initial configuration in a system of three or more members

happens quickly and many will have already produced a hierarchical system by the time

the primary reaches the zero-age main sequence (ZAMS). As stressed above, however, triple

(and higher) systems can still evolve dynamically. This can result in the ejection of a star

– 14 –

(typically the least massive) from the system.

After the ZAMS, since Cepheids are post-red-giant stars, very close binaries will have

undergone Roche-lobe overflow. In extreme cases this results in mergers. Cepheids which

began life as B stars have in some cases had their binary properties altered in this way,

since there are no Cepheid binaries with shorter periods than a year in the Milky Way

(Sugars & Evans 1996)

We stress three features of the current investigation which contribute to the

understanding of binary/multiple properties in fairly massive stars (typically 6M⊙). First,

we have typically searched an area as wide as the canonical 0.1 pc (20,000 AU) in our

HST images for possible companions, which is the extent expected for which physical

companions. Second, the XMM observations provide a severe winnowing of the candidate

list, and an important constraint on the extent of bound companions. Third, the magnitude

difference between the Cepheid and a possible companion can be more than ∆V = 10 mag,

including companions through K stars. A late K star would have a mass ratio of 0.1 with a

6M⊙ Cepheid.

Full discussion of the results of closer companion will be provided in Paper III; however

we will make a preliminary comparison with other recent results, particularly of the extent

of wide systems. The recent survey of resolved O star companions (Sana et al. 2014;

SMASH) plans a further analysis of binary properties; however, a preliminary comparison

with their results is warranted at this stage. The VLT and VLTI instruments they used

cover from 0.5 through 104 mas, (1 AU to 0.1 pc at the typical distance of their O stars

of 2 kpc). (The typical distance in our Cepheid survey is 700 pc). It is only their widest

separations which are comparable to the Cepheids discussed here. An interesting change

occurs for their detections at about 1′′. The magnitude differences are larger (up to 8

mag in H); however while there are many comparatively faint companions, the number

– 15 –

of bright companions drops off as compared with closer regions. Indeed, their probability

analysis also concluded that the low-mass companions at separations ≥2′′ are not physical

companions. For a 20M⊙ system, the separation of ∼2000 AU where this happens is similar

to the widest separation we find.

Tokovinin et al. (2006) made a survey of 165 solar-type spectroscopic binaries to search

for wider components. They noted a decrease in third components at periods >105 years

(≃ 3000 AU). In a more recent distance-limited survey of solar-mass stars, Tokovinin (2014)

finds a decrease in the widest companions at about the same period. A similar result was

found by Raghavan et al. (2010), also for solar-mass stars.

6. Summary

We discuss XMM observations of possible resolved companions of Cepheids from

an HST WFC3 survey. X-ray observations differentiate between young, low-mass stars

(physical companions of Cepheids) and old field stars.

• Of the possible resolved companions, only one was unambiguously detected, implying

that it is a young star. The Cepheid involved is S Nor #4, which is a member of a

well-populated cluster. For this reason, and because the companion is separated from the

Cepheid by as much as 15′′ (13,300 AU), it is likely that this is a cluster member rather

than a gravitationally bound companion.

• Two of the Cepheids (S Mus and R Cru) have an X-ray source which is not resolved

from the Cepheid on the XMM images. However, on a Chandra image of S Mus, the X-rays

are shown to be produced by the Cepheid/spectroscopic binary.

• R Cru and also V659 Cen have additional companions closer than the limit of 5′′ for

this study, which are the likely X-ray sources (to be discussed in Paper III).

– 16 –

• One final X-ray detection (V473 Lyr) has no known companion; hence the prime

suspect is the Cepheid itself. It is a unique Cepheid with a variable amplitude.

• The 14 stars observed with XMM constitute 36% of the candidate companions. None

were found to be young likely physical companions with a separation ≥5′′ or wider than a

projected separation of 4000 AU (assuming that S Nor #4 is a cluster member).

The full binary frequency will be discussed in Paper III including the entire HST

WFC3 survey.

Support for this work was also provided from the Chandra X-ray Center NASA

Contract NAS8-03060. Funding was provided from HST GO-13369.001-A (to NRE and

IP). Vizier and SIMBAD were used in the preparation of this study. Comments from an

anonymous referee improved the clarity of the presentation.

– 17 –

REFERENCES

Aldoretto, E. J., Caballero-Nieves, S. M., Gies, D. R., et al. 2015, AJ, 149, 26

Anderson, R. I., Sahlmann, J., Holl, B., Eyer, L., Palaversa, L., Mowlavi, N., Sueveges, and

Roelens, M. 2015, ApJ, 804, 144

Bohm-Vitense, E., Evans, N. R., Carpenter, K., Beck-Winchatz, B., and Robinson, R. 1997,

ApJ, 477, 916

Burki, G. 1984, IAUS 105, 453

Damiani, F., Maggio, A., Micela, G., and Sciortino, S. 1997a, ApJ, 483, 350

Damiani, F., Maggio, A., Micela, G., and Sciortino, S. 1997b, ApJ,483, 370

Dunstall, P. R., Dufton, P. L,. Sana, H., et al 2015, A&A, 580, 93

Engle, S. G., Guinan, E. F., DePasquale, J., and Evans, N. 2009, AIP Conf Proc., 1135,

192.

Engle, S. G., Guinan,E. F., Harper, G. H., Neilson, H. R. and Evans, N. R. 2014, ApJ 794,

80

Evans, N. R., Massa, D., Fullerton, A., Sonneborn, G., and Iping, R. 2006, AJ, 647, 1387

Evans, N. R., Guinan, E., Engle, S., Wolk, S. J., Schelgel, E., Mason, B. D., Karovska, M.,

and Spitzbart, B. 2010, AJ, 139, 1968

Evans, N. R., Bond, H. E.,Schaefer, G. H. et al. 2013, AJ, 146, 93 (Paper I)

Evans, N. R., Bond, H. E., Schaefer, G. H. et al. 2016a submitted (Paper II)

Evans, N. R., Bond, H. E., Schaefer, G. H. et al. 2016b in preparation (Paper III)

– 18 –

Feigelson, E. D., Getman, K., Townsley, L. 2005, etal. ApJS, 160, 379

Kouwenhoven, M. B. N., Goodwin, S. P., Parker, R. J., Davies, M. B., Malmberg, D., and

Kroupa, P. 2010, MNRAS, 404, 1835

Kraus, A. L., et al. 2012 AJ, 745, 19

Mason, B. D., Gies, D. R., Hartkops, W. I., et al 1998, AJ, 115, 821

Moe, M. and Di Stefano, R. 2015, ApJ, 801, 113

Molnar, L. and Szabados, L. 2014, MNRAS, 442, 3222

Parker, R. J. and Meyer, M. R. 2014, MNRAS, 441, 3722

Perets, H. B. and Kratter, K. M. 2012, ApJ, 760, 99

Pillitteri, I, Micela, G., Sciortino, S. , and Favata, F. 2003, A&A, 399. 919

Pillitteri, I., Evans, N. R., Wolk, S. J. and Syal, M. B. 2013, AJ, 145, 143

Raghavan, D., McAlister, H. A., Henry, T. J., et al. 2010, ApJS, 190, 1

Randich, S., Schmitt, J. H. M. M., Prosser, C. F., and Stauffer, J. R. 1996, A&A, 305, 785

Reipurth, B. and Mikkola, S. 2012, Nature, 492, 221

Ruby, J., Engle, S. G., and Guinan, E. F. 2016, AAS#227

Sana, H. de Mink, S. E., de Koter, A., de Mink, S. E., et al. 2012, Sci., 337, 444

Sana, H., de Koter, A., de Mink, S. E., et al. 2013, A&A, 550, A107

Sana, H., Le Bouquin, J.-B., Lacour, S., et al, 2014, ApJS, 215, 15

Stelzer,B., Flaccomio, E., Montmerle, T. et al. 2005, ApJS, 160, 557

– 19 –

Sugars, B. J. A. and Evans, N. R. 1996, AJ, 112, 1670

Tokovinin, A., Thomas, S., Sterzik, M., and Udry, S. 2006, A&A, 450, 681

Tokovinin, A. 2014, AJ, 147, 87

This manuscript was prepared with the AAS LATEX macros v5.2.

– 20 –

Fig. 1.— a. (left) The XMM image of S Mus. The circles show the location of the Cepheid

(black) and the possible companion (red). The orientation of both figures is the same with N

up and E on the left. The scale is indicated by the 5′′ line. A log stretch is used to emphasize

faint features. Circle sizes in both a and b are arbitrary. b. (right) The HST WFC3 I image

of S Mus. The possible companion is circled.

– 21 –

Fig. 2.— a. (left). The XMM image of R Cru with circles to show the locations of the

Cepheid and two possible companions. b. (right) The HST WFC3 I image of R Cru. Both

possible companions are circled. The image orientation and treatment are the same as for

Figure 1.

– 22 –

Fig. 3.— a. (left). The XMM image of S Nor with circles to show the locations of the

Cepheid and the possible companions. b. (right) The HST WFC3 I image of S Nor. Only

one possible companion was detected in XMM image, the one the furthest to the W. The

image orientation and treatment are the same as for Figure 1.

Fig. 4.— a. (left). The XMM image of V659 Cen with circles to show the locations of the

Cepheid and five possible companions. b. (right) The HST WFC3 I image of V659 Cen. All

possible companions are circled. The image orientation and treatment are the same as for

Figure 1.

– 23 –

Fig. 5.— a. (left). The XMM image of V473 Lyr with circles to show the locations of

the Cepheid and the possible companion. b. (right) The HST WFC3 I image of V473 Lyr.

The possible companion is circled. The image orientation and treatment are the same as for

Figure 1.

– 24 –

0

5×10−3

0.01

norm

aliz

ed c

ount

s s−

1 ke

V−

1

10.5 2

−0.1

0

0.1

0.2

χ

Energy (keV)

Fig. 6.— The spectrum of the S Mus source. Top: normalized counts: +; histogram spectral

fit. Bottom: Differences between data and fit (normalized by the s.d. of the bin).

– 25 –

Fig. 7.— The separations from the Cepheids of the possible companions in Table 2. Those

with possible X-ray detections are hatched. Among the detected sources, S Nor #4 is the star

with a separation of 13300 AU, thought to be a cluster member. The two detected sources

with the smallest separations (S Mus 3950 AU and R Cru 6330 AU) are both thought to

come from closer companions than those in this figure. See text for discussion.

– 26 –

Table 1. XMM Observations

Star Date & UT Exp. E(B − V ) Distance Filter

MOS1 [ks] [pc]

ℓ Car 2010-02-08 10:32:32 52.9 0.17 506 Thick

V659 Cen 2013-09-07 20:10:37 20.8 0.21 753 Medium

V737 Cen 2014-01-26 14:29:20 31.7 0.22 848 Medium

R Cru 2014-01-04 19:56:55 22.0 0.19 829 Medium

S Cru 2013-08-20 21:52:30 11.4 0.16 724 Medium

X Cyg 2013-04-26 02:46:47 30.9 0.29 981 Medium

V473 Lyr 2013-09-22 09:49:34 6.6 0.03 553 Medium

R Mus 2013-02-15 01:19:42 17.4 0.12 844 Medium

S Mus 2013-01-05 14:36:48 25.3 0.21 789 Medium

S Nor 2015-03-13 09:10:05 33.1 0.19 910 Medium

Y Oph 2012-09-12 06:52:49 9.5 0.65 510 Medium

V440 Per 2013-09-02 21:48:18 24.4 0.27 791 Medium

U Sgr 2006-10-11 23:29:34 29.5 0.40 617 Medium

Y Sgr 2013-09-29 04:56:27 13.2 0.20 505 Medium

– 27 –

Table 2. Candidate Resolved Companions

Companion V V − I Sep. PA Sep. R.A. Dec. X-ray

No. [arcsec] [◦] [AU] [J2000] [J2000] Source?

ℓ Car

1 15.21 1.09 19.1 9.7 9,660 09 45 15.3 −62 30 09.5 N

X Cyg

1 18.61 1.68 12.9 96.2 12,700 20 43 24.4 +35 35 28.8 N

2 16.33 1.36 14.8 298.3 14,500 20 43 23.5 +35 35 03.9 N

V659 Cen

1 15.87 1.20 22.1 335.3 16,600 13 31 32.1 −61 34 36.3 N

2 18.09 1.59 14.7 340.7 11,100 13 31 32.7 −61 34 42.6 N

3 16.57 1.27 20.2 238.5 15,200 13 31 30.9 −61 35 07.1 N

4 15.73 1.17 23.8 81.1 17,900 13 31 36.7 −61 34 52.8 N

5 17.59 1.47 17.0 58.6 12,800 13 31 35.4 −61 34 47.6 N

V737 Cen

1 17.22 1.61 7.3 294.9 6,190 14 37 11.0 −62 00 36.1 N

2 17.67 1.61 17.1 231.1 14,500 14 37 10.0 −62 00 49.9 N

R Cru

1 16.28 1.17 7.64 301.7 6,330 12 23 36.7 −61 37 41.1 ?

2 17.94 1.45 23.92 199.5 19,800 12 23 36.5 -61 38 07.8 N

S Cru

1 16.59 1.58 13.8 70.4 9,990 12 54 23.6 −58 25 45.5 N

2 17.90 1.55 11.9 20.0 8,620 12 54 22.5 −58 25 39.0 N

– 28 –

Table 2—Continued

Companion V V − I Sep. PA Sep. R.A. Dec. X-ray

No. [arcsec] [◦] [AU] [J2000] [J2000] Source?

V473 Lyr

1 14.89 1.23 15.0 44.1 8,300 19 16 00.3 +27 55 45.2 N

R Mus

1 15.68 1.17 6.9 328.1 5,820 12 42 04.3 −69 24 21.5 N

S Mus

1 17.94 1.56 5.0 182.5 3,950 12 12 46.9 −70 09 11.3 ?

S Nor

1 16.45 1.15 19.8 172.0 18,000 16 18 52.2 −57 54 19.3 N

2 16.32 1.20 20.1 188.5 18,300 16 18 51.5 −57 54 19.4 N

3 18.06 1.44 8.5 44.1 7,740 16 18 52.6 −57 53 53.5 N

4 13.95 0.90 14.6 288.9 13,300 16 18 50.1 −57 53 54.9 Y

5 18.00 1.52 15.0 1.1 13,600 16 18 51.9 −57 53 44.6 N

6 17.37 1.54 13.5 6.6 12,300 16 18 52.0 −57 53 46.2 N

Y Oph

1 17.13 2.00 18.1 211.9 9,230 17 52 38.1 −06 08 52.5 N

V440

1 15.72 1.21 10.9 305.2 8,620 02 23 50.7 +55 21 59.7 N

2 13.83 1.01 10.6 130.6 8,380 02 23 52.7 +55 21 46.5 N

U Sgr

1 16.31 1.52 19.4 18.1 12,000 18 31 53.8 −19 07 11.8 N

– 29 –

Table 2—Continued

Companion V V − I Sep. PA Sep. R.A. Dec. X-ray

No. [arcsec] [◦] [AU] [J2000] [J2000] Source?

2 17.42 1.95 13.9 126.7 8,580 18 31 54.1 −19 07 38.5 N

3 17.66 1.94 17.1 163.8 10,500 18 31 53.7 −19 07 46.7 N

Y Sgr

1 17.06 1.85 10.6 204.2 5,350 18 21 22.7 −18 51 45.7 N

– 30 –

Table 3. XMM Upper Limits of Candidate Companions

Cepheid Companion Upper Limit NH Flux logLX

No. [ct ks−1] [1021 cm−2] [10−15 ergs cm−2 s−1] [ergs s−1]

ℓ Car 1 0.197 1.00 1.72 28.72

X Cyg 1 0.739 1.71 6.66 29.89

2 0.797 7.18 29.92

V659 Cen 1 0.230 1.24 1.82 29.09

2 0.238 1.89 29.11

3 0.234 1.86 29.10

4 0.229 1.82 29.09

5 0.234 1.86 29.10

V737 Cen 1 0.358 1.30 2.89 29.40

2 0.369 2.98 29.41

R Cru 2 0.428 1.12 3.28 29.43

S Cru 1 0.876 0.94 6.38 29.60

2 0.867 6.31 29.60

V473 Lyr 2.35 0.18 16.1 29.76

R Mus 1 0.514 0.71 3.49 29.48

S Nor 1 0.340 1.12 2.61 29.41

2 0.341 2.62 29.42

3 0.359 2.75 29.44

5 0.375 2.88 29.47

6 0.374 2.87 29.46

– 31 –

Table 3—Continued

Cepheid Companion Upper Limit NH Flux logLX

No. [ct ks−1] [1021 cm−2] [10−15 ergs cm−2 s−1] [ergs s−1]

Y Oph 1 0.420 3.83 6.31 29.29

V440 Per 1 0.231 1.49 1.97 29.17

2 0.223 1.90 29.15

U Sgr 1 0.238 2.36 2.53 29.06

2 0.229 2.44 29.05

3 0.231 2.46 29.05

Y Sgr 1 0.747 1.18 5.83 29.25

Table 4. XMM Detections

Star Source NH Flux logLX Med E Comments

[ct ks−1] [1021 cm−2] [10−15 ergs cm−2 s−1] [ergs s−1] [keV]

V659 Cen 5.97 1.24 4.74 29.51 0.88 Cep? 0.′′7 comp

R Cru 1.00 1.12 7.67 29.80 1.24 Cep? 1.′′9 comp

V473 Lyr 3.61 0.18 20.5 29.88 0.66 Cep?

S Mus 4.36 1.24 34.6 30.46 0.93 Cep?

S Nor #4 1.32 1.12 10.1 30.00 0.95